This article first appeared in our sister publication, Dairy Farmer.
Rapid advancements in technology are revolutionising decision-making on farms. Artificial intelligence (AI) and machine learning are becoming increasingly common, allowing dairy farmers to utilise intelligent algorithms to drive efficiency in their operations. By identifying potential problems early, improving production and increasing profitability, AI and machine learning are game-changers that will take farming to the next level.
“Farmers need tools to be able to make the right decisions as there are tougher consequences from not getting it right nowadays,” OmniEye chief executive officer Andrew Christie says.
“Farming is becoming increasingly competitive with more pressure on margins and everyone is trying to remain profitable and looking for opportunities to improve efficiency.”
The locomotion monitoring system developed by OmniEye is a prime example of AI supporting farmers to make better decisions based on data. The significant economic and welfare impacts lameness can have on a farming business are what intrigued OmniEye founders Greg Peyroux and Benoit Auvray. They had been looking for a problem to solve.
“During the first lockdown Greg and Benoit had spent hundreds of hours on the phone to farmers and people throughout the sector exploring what challenges farmers were facing,” Christie says.
“They both have agritech backgrounds and were looking for a way to utilise a camera, machine learning and AI to support farmers.
“Lameness was a clear winner because of the extensive impacts it has on the farming business and for the animal involved.”
They got stuck in, developing a prototype, using veterinarians to train the model while working closely with Pāmu and other pilot farms to validate and improve it. The scoring is based on the DairyNZ lameness scoring scale.
“We have over 30,000 vet scores, which we used to train the system to identify a healthy cow or a lame cow and everything in between.
“So when it sees a cow, it assigns a score based on the DairyNZ scale and presents the data on a dashboard for the farmer, who can set thresholds for alerts and make decisions based on real-time data.”
The cow walks past an electronic ID tag reader at the exit race and when she comes into the view of the camera it identifies and tracks her, then assigns a score after analysing her gait from machine learning.
Christie can see an increasing need for innovative tools to assist decision-making.
“A tool like ours can help farmers monitor what is happening with their animals, help them identify problems early and look for patterns or trends and monitor progress over time.
“They can also monitor cows that are undergoing treatment to help regain levels of productivity that problems like lameness have a real impact on.”
He believes farmers will turn to AI-based tools more and more as they seek detailed information to make decisions. But the farm system and individual situation will influence which tools each farm adopts, as well as the return on investment they can expect to see.
“Farmers will need to identify where the biggest bang for their buck is, which starts by identifying how much an issue is impacting their productivity, and then working through how the technology could improve the outcomes and how they make their decisions surrounding it.”
With over 30 systems throughout New Zealand already, he knows the appetite for technology is growing but believes we are only at the start of the AI and machine learning revolution. And our understanding of what the emerging technology is capable of will only improve with time.
“It’s an incredibly exciting space to be. There are so many complex problems on farms but that also means there are so many opportunities for innovation and continued technology development.
“And New Zealand agriculture is already good at working together to help everyone improve, so more data and tools being available will certainly drive the efficiencies and improved decision-making that we’re all searching for.”
